Unipolar and adaptive organic synaptic devices with resilient plasticity for artificial neuromorphic intelligence

Abstract

Abstract Flexible neuromorphic electronic devices hold great promise for wearable computing and soft robotics in the artificial intelligence era, requiring inspiration from biological neural systems to achieve adaptive and efficient signal processing. The quest for devices with resilient synaptic plasticity that fully replicate the functions of biological counterparts is paramount for realising dynamically reconfigurable neuromorphic architectures and overcoming the limitations of conventional electronics. Here, we present an artificial synaptic device based on the organic small molecule vanadyl phthalocyanine, which exploits charge trapping to enable synaptic weight modulation beyond a single form of plasticity. The device exhibits both synaptic potentiation and depression under voltages bias of the same polarity but different amplitudes. It also demonstrates a sustained response to presynaptic stimuli while allowing controlled signal attenuation for adaptability across complex environments. Furthermore, the device emulates key biological nociceptive functions, including self-protection mechanisms, by leveraging its unique electrical characteristics. It maintains excellent thermal adaptability up to 373 K and exhibits outstanding mechanical flexibility, making it highly suitable for wearable health-monitoring electronics with sensory adaptability. With minimal reconfiguration, this work aligns with the evolving demands of neuromorphic computing, offering a promising pathway towards bridging artificial intelligence and hardware implementation.